In the thread regarding breakpoints on "if" statements, it finally became clear that traditional debugger features like "Step Into" and "Step Over" are simply not applicable to an injected code mechanism.  However, adding a "Pre-Pend" option might provide a workaround for certain cases.

In the referenced tread it was explained that if the original program statement is

    abc = 5 ;

then the injected breakpoint code would follow the original, i.e.

   abc = 5 ; Serial.println( "some debug notification" ) ;

How difficult would it be to implement a Pre-Pend option to breakpoint definitions such that the code sent to the compiler would look like

    Serial.println( "some debug notification" ) ; abc = 5 ;

This gets around part of the problem with setting  break on an "if" statement that has a condition which will test false.  By adding suitable conditionals to the breakpoint I could at least signal whether the statement body will execute.  It seems to me this is superior to having the appended trace only execute when the "if" condition is true.

Hi.

As far as I understand, it is very similar to send debug information before or after the execution of the current line, as it is just matter of where you locate the break point:

If you need to stop and evaluate the expresion before if condition, you can just mark the break point the line before with debug information about the condition.

Regards.

Fair enough.  Certainly good enough for V1.